def export_result_dymola(model,
data,
file_name='',
format='txt',
scaled=False):
"""
Export an optimization or simulation result to file in Dymolas result file
format. The parameter values are read from the z vector of the model object
and the time series are read from the data argument.
Parameters::
model --
A Model object.
data --
A two dimensional array of variable trajectory data. The first
column represents the time vector. The following colums contain, in
order, the derivatives, the states, the inputs and the algebraic
variables. The ordering is according to increasing value references.
file_name --
If no file name is given, the name of the model (as defined by
FMUModel.get_identifier()) concatenated with the string '_result' is used.
A file suffix equal to the format argument is then appended to the
file name.
Default: Empty string.
format --
A text string equal either to 'txt' for textual format or 'mat' for
binary Matlab format.
Default: 'txt'
scaled --
Set this parameter to True to write the result to file without
taking scaling into account. If the value of scaled is False, then
the variable scaling factors of the model are used to reproduced the
unscaled variable values.
Default: False
Limitations::
Currently only textual format is supported.
"""
if (format == 'txt'):
if file_name == '':
file_name = model.get_identifier() + '_result.txt'
# Open file
f = codecs.open(file_name, 'w', 'utf-8')
# Write header
f.write('#1\n')
f.write('char Aclass(3,11)\n')
f.write('Atrajectory\n')
f.write('1.1\n')
f.write('\n')
md = model._get_XMLDoc()
# NOTE: it is essential that the lists 'names', 'aliases', 'descriptions'
# and 'variabilities' are sorted in the same order and that this order
# is: value reference order AND within the same value reference the
# non-alias variable must be before its corresponding aliases. Otherwise
# the header-writing algorithm further down will fail.
# Therefore the following code is needed...
# all lists that we need for later
vrefs_alias = []
vrefs = []
names_alias = []
names = []
aliases_alias = []
aliases = []
descriptions_alias = []
descriptions = []
variabilities_alias = []
variabilities = []
# go through all variables and split in non-alias/only-alias lists
for var in md.get_model_variables():
if var.get_alias() == xmlparser.NO_ALIAS:
vrefs.append(var.get_value_reference())
names.append(var.get_name())
aliases.append(var.get_alias())
descriptions.append(var.get_description())
variabilities.append(var.get_variability())
else:
vrefs_alias.append(var.get_value_reference())
names_alias.append(var.get_name())
aliases_alias.append(var.get_alias())
descriptions_alias.append(var.get_description())
variabilities_alias.append(var.get_variability())
# extend non-alias lists with only-alias-lists
vrefs.extend(vrefs_alias)
names.extend(names_alias)
aliases.extend(aliases_alias)
descriptions.extend(descriptions_alias)
variabilities.extend(variabilities_alias)
# zip to list of tuples and sort - non alias variables are now
# guaranteed to be first in list and all variables are in value reference
# order
names = sorted(zip(tuple(vrefs), tuple(names)), key=itemgetter(0))
aliases = sorted(zip(tuple(vrefs), tuple(aliases)), key=itemgetter(0))
descriptions = sorted(zip(tuple(vrefs), tuple(descriptions)),
key=itemgetter(0))
variabilities = sorted(zip(tuple(vrefs), tuple(variabilities)),
key=itemgetter(0))
num_vars = len(names)
# Find the maximum name and description length
max_name_length = len('Time')
max_desc_length = len('Time in [s]')
for i in range(len(names)):
name = names[i][1]
desc = descriptions[i][1]
if (len(name) > max_name_length):
max_name_length = len(name)
if (len(desc) > max_desc_length):
max_desc_length = len(desc)
f.write('char name(%d,%d)\n' % (num_vars + 1, max_name_length))
f.write('time\n')
# write names
for name in names:
f.write(name[1] + '\n')
f.write('\n')
f.write('char description(%d,%d)\n' % (num_vars + 1, max_desc_length))
f.write('Time in [s]\n')
# write descriptions
for desc in descriptions:
f.write(desc[1] + '\n')
f.write('\n')
# Write data meta information
offs = model.get_offsets()
n_parameters = offs[12] # offs[12] = offs_dx
f.write('int dataInfo(%d,%d)\n' % (num_vars + 1, 4))
f.write('0 1 0 -1 # time\n')
cnt_1 = 1
cnt_2 = 1
for i, name in enumerate(names):
(ref, type) = jmi._translate_value_ref(name[0])
if int(ref) < n_parameters: # Put parameters in data set
if aliases[i][1] == 0: # no alias
cnt_1 = cnt_1 + 1
f.write('1 %d 0 -1 # ' % cnt_1 + name[1] + '\n')
elif aliases[i][1] == 1: # alias
f.write('1 %d 0 -1 # ' % cnt_1 + name[1] + '\n')
else: # negated alias
f.write('1 -%d 0 -1 # ' % cnt_1 + name[1] + '\n')
else:
if aliases[i][1] == 0: # noalias
cnt_2 = cnt_2 + 1
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] + '\n')
elif aliases[i][1] == 1: # alias
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] + '\n')
else: #neg alias
f.write('2 -%d 0 -1 # ' % cnt_2 + name[1] + '\n')
f.write('\n')
sc = model.jmimodel.get_variable_scaling_factors()
z = model.z
rescale = (model.get_scaling_method()
== jmi.JMI_SCALING_VARIABLES) and (not scaled)
# Write data
# Write data set 1
f.write('float data_1(%d,%d)\n' % (2, n_parameters + 1))
f.write("%.14E" % data[0, 0])
str_text = ''
for ref in range(n_parameters):
#print ref
if rescale:
#print z[ref]*sc[ref]
#print "hej"
str_text += " %.14E" % (z[ref] * sc[ref])
else:
#print z[ref]
#print "hopp"
str_text += " %.14E" % (z[ref])
f.write(str_text)
f.write('\n')
f.write("%.14E" % data[-1, 0])
f.write(str_text)
f.write('\n\n')
# Write data set 2
n_vars = len(data[0, :])
n_points = len(data[:, 0])
f.write('float data_2(%d,%d)\n' % (n_points, n_vars))
for i in range(n_points):
str = ''
for ref in range(n_vars):
if ref == 0: # Don't scale time
str = str + (" %.14E" % data[i, ref])
else:
if rescale:
str = str + (
" %.14E" %
(data[i, ref] * sc[ref - 1 + n_parameters]))
else:
str = str + (" %.14E" % data[i, ref])
f.write(str + '\n')
f.write('\n')
f.close()
else:
raise Error('Export on binary Dymola result files not yet supported.')
def write_header(self, file_name='', scaled=False):
"""
Opens the file and writes the header. This includes the
information about the variables and a table determining the link
between variables and data.
Parameters::
file_name --
If no file name is given, the name of the model (as
defined by JMUModel.get_identifier()) concatenated with the
string '_result' is used. A file suffix equal to the
format argument is then appended to the file name.
Default: Empty string.
"""
model = self.model
if file_name=='':
file_name=model.get_identifier() + '_result.txt'
# Open file
f = codecs.open(file_name,'w','utf-8')
self._file_open = True
# Write header
f.write('#1\n')
f.write('char Aclass(3,11)\n')
f.write('Atrajectory\n')
f.write('1.1\n')
f.write('\n')
# Retrieve the xml-file
md = model._get_XMLDoc()
# all lists that we need for later
vrefs_alias = []
vrefs = []
names_alias = []
names = []
aliases_alias = []
aliases = []
descriptions_alias = []
descriptions = []
variabilities_alias = []
variabilities = []
# go through all variables and split in non-alias/only-alias lists
for var in md.get_model_variables():
if var.get_alias() == xmlparser.NO_ALIAS:
vrefs.append(var.get_value_reference())
names.append(var.get_name())
aliases.append(var.get_alias())
descriptions.append(var.get_description())
variabilities.append(var.get_variability())
else:
vrefs_alias.append(var.get_value_reference())
names_alias.append(var.get_name())
aliases_alias.append(var.get_alias())
descriptions_alias.append(var.get_description())
variabilities_alias.append(var.get_variability())
# Parameters for sensitivity calculations
sens_p = model.get_p_opt_variable_names()
sens_x = model.get_x_variable_names()
sens_w = model.get_w_variable_names()
# sort in value reference order (must match order in data)
sens_xw = sens_x+sens_w
sens_xw = sorted(sens_xw, key=itemgetter(0))
sens_names = []
sens_desc = []
for j in range(len(sens_p)):
#(ref_param, type) = pyjmi.jmi._translate_value_ref(sens_p[j][0])
for i in range(len(sens_xw)):
sens_names += ['d'+sens_xw[i][1]+'/d'+sens_p[j][1]]
sens_desc += ['Sensitivity of '+sens_xw[i][1]+' with respect to '+sens_p[j][1]+'.']
# extend non-alias lists with only-alias-lists
vrefs.extend(vrefs_alias)
names.extend(names_alias)
aliases.extend(aliases_alias)
descriptions.extend(descriptions_alias)
variabilities.extend(variabilities_alias)
# zip to list of tuples and sort - non alias variables are now
# guaranteed to be first in list and all variables are in value reference
# order
names = sorted(zip(
tuple(vrefs),
tuple(names)),
key=itemgetter(0))
aliases = sorted(zip(
tuple(vrefs),
tuple(aliases)),
key=itemgetter(0))
descriptions = sorted(zip(
tuple(vrefs),
tuple(descriptions)),
key=itemgetter(0))
variabilities = sorted(zip(
tuple(vrefs),
tuple(variabilities)),
key=itemgetter(0))
# sort in value reference order (must match order in data)
#names = sorted(md.get_variable_names(), key=itemgetter(0))
#aliases = sorted(md.get_variable_aliases(), key=itemgetter(0))
#descriptions = sorted(md.get_variable_descriptions(), key=itemgetter(0))
#variabilities = sorted(md.get_variable_variabilities(), key=itemgetter(0))
num_vars = len(names)+len(sens_names)
# Find the maximum name and description length
max_name_length = len('Time')
max_desc_length = len('Time in [s]')
for i in range(len(names)):
name = names[i][1]
desc = descriptions[i][1]
if (len(name)>max_name_length):
max_name_length = len(name)
if (len(desc)>max_desc_length):
max_desc_length = len(desc)
for i in range(len(sens_names)):
if (len(sens_names[i])>max_name_length):
max_name_length = len(sens_names[i])
if (len(sens_desc[i])>max_desc_length):
max_desc_length = len(sens_desc[i])
f.write('char name(%d,%d)\n' % (num_vars + 1, max_name_length))
f.write('time\n')
self._rescale = (model.get_scaling_method() == jmi.JMI_SCALING_VARIABLES) and (not scaled)
for name in names:
f.write(name[1] +'\n')
for name in sens_names:
f.write(name+'\n')
f.write('\n')
# Write descriptions
f.write('char description(%d,%d)\n' % (num_vars + 1, max_desc_length))
f.write('Time in [s]\n')
# write descriptions
for desc in descriptions:
f.write(desc[1]+'\n')
# Write sensitivity descriptions
for desc in sens_desc:
f.write(desc+'\n')
f.write('\n')
# Write data meta information
offs = model.get_offsets()
n_parameters = offs[12] # offs[12] = offs_dx
self._n_parameters = n_parameters
f.write('int dataInfo(%d,%d)\n' % (num_vars + 1, 4))
f.write('0 1 0 -1 # time\n')
cnt_1 = 1
cnt_2 = 1
for i, name in enumerate(names):
(ref, type) = jmi._translate_value_ref(name[0])
if int(ref) < n_parameters: # Put parameters in data set
if aliases[i][1] == 0: # no alias
cnt_1 = cnt_1 + 1
f.write('1 %d 0 -1 # ' % cnt_1 + name[1]+'\n')
elif aliases[i][1] == 1: # alias
f.write('1 %d 0 -1 # ' % cnt_1 + name[1]+'\n')
else: # negated alias
f.write('1 -%d 0 -1 # ' % cnt_1 + name[1] +'\n')
else:
if aliases[i][1] == 0: # noalias
cnt_2 = cnt_2 + 1
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] +'\n')
elif aliases[i][1] == 1: # alias
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] +'\n')
else: #neg alias
f.write('2 -%d 0 -1 # ' % cnt_2 + name[1] +'\n')
self._nvariables_without_sens = cnt_2
sc = model.jmimodel.get_variable_scaling_factors()
self._sc = sc
z = model.z
sens_param_res = []
self._sens_sc = [] #Sensitivity scaling factors
#Write sensitivity variables into the table (No alias, no parameters)
for i,name in enumerate(sens_names):
name_split = name.split('/')
param = name_split[1][1:]
var = name_split[0][1:]
(ref_param, type) = jmi._translate_value_ref(model.get_value_reference(param))
(ref_var , type) = jmi._translate_value_ref(model.get_value_reference(var))
if int(ref_var) < n_parameters: # Put parameters in data set
cnt_1 = cnt_1 + 1
f.write('1 %d 0 -1 # ' % cnt_1 + name +'\n')
if ref_param == ref_var:
if model.is_negated_alias(var):
sens_param_res += [-1.0]
else:
sens_param_res += [1.0]
else:
sens_param_res += [0.0]
else:
if not md.is_alias(var):
cnt_2 = cnt_2+1
self._sens_sc += [self._sc[ref_var]/self._sc[ref_param]]
f.write('2 %d 0 -1 # ' % cnt_2 + name + '\n')
elif md.is_negated_alias(var):
f.write('2 -%d 0 -1 # ' % cnt_2 + name + '\n')
else:
f.write('2 %d 0 -1 # ' % cnt_2 + name + '\n')
self._nvariables_total = cnt_2 #Store the number of variables
f.write('\n')
# Write data
# Write data set 1
f.write('float data_1(%d,%d)\n' % (2, cnt_1))
str_text = ''
for ref in range(n_parameters):
if self._rescale:
str_text += " %.14E" % (z[ref]*sc[ref])
else:
str_text += " %.14E" % (z[ref])
# Write sensitivity data set 1
for i in range(cnt_1-1-n_parameters):
str_text += " %.14E " % sens_param_res[i]
#f.write("%.14E" % data[0,0])
self._point_first_t = f.tell()
f.write("%s" % ' '*28)
f.write(str_text)
f.write('\n')
self._point_last_t = f.tell()
f.write("%s" % ' '*28)
f.write(str_text)
f.write('\n\n')
f.write('float data_2(')
self._point_npoints = f.tell()
f.write(' '*(14+4+14))
f.write('\n')
self._file = f
def write_header(self, file_name='', scaled=False):
"""
Opens the file and writes the header. This includes the
information about the variables and a table determining the link
between variables and data.
Parameters::
file_name --
If no file name is given, the name of the model (as
defined by FMUModel.get_identifier()) concatenated with the
string '_result' is used. A file suffix equal to the
format argument is then appended to the file name.
Default: Empty string.
"""
model = self.model
if file_name == '':
file_name = model.get_identifier() + '_result.txt'
# Open file
f = codecs.open(file_name, 'w', 'utf-8')
self._file_open = True
# Write header
f.write('#1\n')
f.write('char Aclass(3,11)\n')
f.write('Atrajectory\n')
f.write('1.1\n')
f.write('\n')
# Retrieve the xml-file
md = model._get_XMLDoc()
# all lists that we need for later
vrefs_alias = []
vrefs = []
names_alias = []
names = []
aliases_alias = []
aliases = []
descriptions_alias = []
descriptions = []
variabilities_alias = []
variabilities = []
# go through all variables and split in non-alias/only-alias lists
for var in md.get_model_variables():
if var.get_alias() == xmlparser.NO_ALIAS:
vrefs.append(var.get_value_reference())
names.append(var.get_name())
aliases.append(var.get_alias())
descriptions.append(var.get_description())
variabilities.append(var.get_variability())
else:
vrefs_alias.append(var.get_value_reference())
names_alias.append(var.get_name())
aliases_alias.append(var.get_alias())
descriptions_alias.append(var.get_description())
variabilities_alias.append(var.get_variability())
# Parameters for sensitivity calculations
sens_p = model.get_p_opt_variable_names()
sens_x = model.get_x_variable_names()
sens_w = model.get_w_variable_names()
# sort in value reference order (must match order in data)
sens_xw = sens_x + sens_w
sens_xw = sorted(sens_xw, key=itemgetter(0))
sens_names = []
sens_desc = []
for j in range(len(sens_p)):
#(ref_param, type) = pyjmi.jmi._translate_value_ref(sens_p[j][0])
for i in range(len(sens_xw)):
sens_names += ['d' + sens_xw[i][1] + '/d' + sens_p[j][1]]
sens_desc += [
'Sensitivity of ' + sens_xw[i][1] + ' with respect to ' +
sens_p[j][1] + '.'
]
# extend non-alias lists with only-alias-lists
vrefs.extend(vrefs_alias)
names.extend(names_alias)
aliases.extend(aliases_alias)
descriptions.extend(descriptions_alias)
variabilities.extend(variabilities_alias)
# zip to list of tuples and sort - non alias variables are now
# guaranteed to be first in list and all variables are in value reference
# order
names = sorted(zip(tuple(vrefs), tuple(names)), key=itemgetter(0))
aliases = sorted(zip(tuple(vrefs), tuple(aliases)), key=itemgetter(0))
descriptions = sorted(zip(tuple(vrefs), tuple(descriptions)),
key=itemgetter(0))
variabilities = sorted(zip(tuple(vrefs), tuple(variabilities)),
key=itemgetter(0))
# sort in value reference order (must match order in data)
#names = sorted(md.get_variable_names(), key=itemgetter(0))
#aliases = sorted(md.get_variable_aliases(), key=itemgetter(0))
#descriptions = sorted(md.get_variable_descriptions(), key=itemgetter(0))
#variabilities = sorted(md.get_variable_variabilities(), key=itemgetter(0))
num_vars = len(names) + len(sens_names)
# Find the maximum name and description length
max_name_length = len('Time')
max_desc_length = len('Time in [s]')
for i in range(len(names)):
name = names[i][1]
desc = descriptions[i][1]
if (len(name) > max_name_length):
max_name_length = len(name)
if (len(desc) > max_desc_length):
max_desc_length = len(desc)
for i in range(len(sens_names)):
if (len(sens_names[i]) > max_name_length):
max_name_length = len(sens_names[i])
if (len(sens_desc[i]) > max_desc_length):
max_desc_length = len(sens_desc[i])
f.write('char name(%d,%d)\n' % (num_vars + 1, max_name_length))
f.write('time\n')
self._rescale = (model.get_scaling_method()
== jmi.JMI_SCALING_VARIABLES) and (not scaled)
for name in names:
f.write(name[1] + '\n')
for name in sens_names:
f.write(name + '\n')
f.write('\n')
# Write descriptions
f.write('char description(%d,%d)\n' % (num_vars + 1, max_desc_length))
f.write('Time in [s]\n')
# write descriptions
for desc in descriptions:
f.write(desc[1] + '\n')
# Write sensitivity descriptions
for desc in sens_desc:
f.write(desc + '\n')
f.write('\n')
# Write data meta information
offs = model.get_offsets()
n_parameters = offs[12] # offs[12] = offs_dx
self._n_parameters = n_parameters
f.write('int dataInfo(%d,%d)\n' % (num_vars + 1, 4))
f.write('0 1 0 -1 # time\n')
cnt_1 = 1
cnt_2 = 1
for i, name in enumerate(names):
(ref, type) = jmi._translate_value_ref(name[0])
if int(ref) < n_parameters: # Put parameters in data set
if aliases[i][1] == 0: # no alias
cnt_1 = cnt_1 + 1
f.write('1 %d 0 -1 # ' % cnt_1 + name[1] + '\n')
elif aliases[i][1] == 1: # alias
f.write('1 %d 0 -1 # ' % cnt_1 + name[1] + '\n')
else: # negated alias
f.write('1 -%d 0 -1 # ' % cnt_1 + name[1] + '\n')
else:
if aliases[i][1] == 0: # noalias
cnt_2 = cnt_2 + 1
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] + '\n')
elif aliases[i][1] == 1: # alias
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] + '\n')
else: #neg alias
f.write('2 -%d 0 -1 # ' % cnt_2 + name[1] + '\n')
self._nvariables_without_sens = cnt_2
sc = model.jmimodel.get_variable_scaling_factors()
self._sc = sc
z = model.z
sens_param_res = []
self._sens_sc = [] #Sensitivity scaling factors
#Write sensitivity variables into the table (No alias, no parameters)
for i, name in enumerate(sens_names):
name_split = name.split('/')
param = name_split[1][1:]
var = name_split[0][1:]
(ref_param,
type) = jmi._translate_value_ref(model.get_value_reference(param))
(ref_var,
type) = jmi._translate_value_ref(model.get_value_reference(var))
if int(ref_var) < n_parameters: # Put parameters in data set
cnt_1 = cnt_1 + 1
f.write('1 %d 0 -1 # ' % cnt_1 + name + '\n')
if ref_param == ref_var:
if model.is_negated_alias(var):
sens_param_res += [-1.0]
else:
sens_param_res += [1.0]
else:
sens_param_res += [0.0]
else:
if not md.is_alias(var):
cnt_2 = cnt_2 + 1
self._sens_sc += [self._sc[ref_var] / self._sc[ref_param]]
f.write('2 %d 0 -1 # ' % cnt_2 + name + '\n')
elif md.is_negated_alias(var):
f.write('2 -%d 0 -1 # ' % cnt_2 + name + '\n')
else:
f.write('2 %d 0 -1 # ' % cnt_2 + name + '\n')
self._nvariables_total = cnt_2 #Store the number of variables
f.write('\n')
# Write data
# Write data set 1
f.write('float data_1(%d,%d)\n' % (2, cnt_1))
str_text = ''
for ref in range(n_parameters):
if self._rescale:
str_text += " %.14E" % (z[ref] * sc[ref])
else:
str_text += " %.14E" % (z[ref])
# Write sensitivity data set 1
for i in range(cnt_1 - 1 - n_parameters):
str_text += " %.14E " % sens_param_res[i]
#f.write("%.14E" % data[0,0])
self._point_first_t = f.tell()
f.write("%s" % ' ' * 28)
f.write(str_text)
f.write('\n')
self._point_last_t = f.tell()
f.write("%s" % ' ' * 28)
f.write(str_text)
f.write('\n\n')
f.write('float data_2(')
self._point_npoints = f.tell()
f.write(' ' * (14 + 4 + 14))
f.write('\n')
self._file = f
def export_result_dymola(model, data, file_name='', format='txt', scaled=False):
"""
Export an optimization or simulation result to file in Dymolas result file
format. The parameter values are read from the z vector of the model object
and the time series are read from the data argument.
Parameters::
model --
A Model object.
data --
A two dimensional array of variable trajectory data. The first
column represents the time vector. The following colums contain, in
order, the derivatives, the states, the inputs and the algebraic
variables. The ordering is according to increasing value references.
file_name --
If no file name is given, the name of the model (as defined by
JMUModel.get_identifier()) concatenated with the string '_result' is used.
A file suffix equal to the format argument is then appended to the
file name.
Default: Empty string.
format --
A text string equal either to 'txt' for textual format or 'mat' for
binary Matlab format.
Default: 'txt'
scaled --
Set this parameter to True to write the result to file without
taking scaling into account. If the value of scaled is False, then
the variable scaling factors of the model are used to reproduced the
unscaled variable values.
Default: False
Limitations::
Currently only textual format is supported.
"""
if (format=='txt'):
if file_name=='':
file_name=model.get_identifier() + '_result.txt'
# Open file
f = codecs.open(file_name,'w','utf-8')
# Write header
f.write('#1\n')
f.write('char Aclass(3,11)\n')
f.write('Atrajectory\n')
f.write('1.1\n')
f.write('\n')
md = model._get_XMLDoc()
# NOTE: it is essential that the lists 'names', 'aliases', 'descriptions'
# and 'variabilities' are sorted in the same order and that this order
# is: value reference order AND within the same value reference the
# non-alias variable must be before its corresponding aliases. Otherwise
# the header-writing algorithm further down will fail.
# Therefore the following code is needed...
# all lists that we need for later
vrefs_alias = []
vrefs = []
names_alias = []
names = []
aliases_alias = []
aliases = []
descriptions_alias = []
descriptions = []
variabilities_alias = []
variabilities = []
# go through all variables and split in non-alias/only-alias lists
for var in md.get_model_variables():
if var.get_alias() == xmlparser.NO_ALIAS:
vrefs.append(var.get_value_reference())
names.append(var.get_name())
aliases.append(var.get_alias())
descriptions.append(var.get_description())
variabilities.append(var.get_variability())
else:
vrefs_alias.append(var.get_value_reference())
names_alias.append(var.get_name())
aliases_alias.append(var.get_alias())
descriptions_alias.append(var.get_description())
variabilities_alias.append(var.get_variability())
# extend non-alias lists with only-alias-lists
vrefs.extend(vrefs_alias)
names.extend(names_alias)
aliases.extend(aliases_alias)
descriptions.extend(descriptions_alias)
variabilities.extend(variabilities_alias)
# zip to list of tuples and sort - non alias variables are now
# guaranteed to be first in list and all variables are in value reference
# order
names = sorted(zip(
tuple(vrefs),
tuple(names)),
key=itemgetter(0))
aliases = sorted(zip(
tuple(vrefs),
tuple(aliases)),
key=itemgetter(0))
descriptions = sorted(zip(
tuple(vrefs),
tuple(descriptions)),
key=itemgetter(0))
variabilities = sorted(zip(
tuple(vrefs),
tuple(variabilities)),
key=itemgetter(0))
num_vars = len(names)
# Find the maximum name and description length
max_name_length = len('Time')
max_desc_length = len('Time in [s]')
for i in range(len(names)):
name = names[i][1]
desc = descriptions[i][1]
if (len(name)>max_name_length):
max_name_length = len(name)
if (len(desc)>max_desc_length):
max_desc_length = len(desc)
f.write('char name(%d,%d)\n' % (num_vars + 1, max_name_length))
f.write('time\n')
# write names
for name in names:
f.write(name[1] +'\n')
f.write('\n')
f.write('char description(%d,%d)\n' % (num_vars + 1, max_desc_length))
f.write('Time in [s]\n')
# write descriptions
for desc in descriptions:
f.write(desc[1]+'\n')
f.write('\n')
# Write data meta information
offs = model.get_offsets()
n_parameters = offs[12] # offs[12] = offs_dx
f.write('int dataInfo(%d,%d)\n' % (num_vars + 1, 4))
f.write('0 1 0 -1 # time\n')
cnt_1 = 1
cnt_2 = 1
for i, name in enumerate(names):
(ref, type) = jmi._translate_value_ref(name[0])
if int(ref) < n_parameters: # Put parameters in data set
if aliases[i][1] == 0: # no alias
cnt_1 = cnt_1 + 1
f.write('1 %d 0 -1 # ' % cnt_1 + name[1]+'\n')
elif aliases[i][1] == 1: # alias
f.write('1 %d 0 -1 # ' % cnt_1 + name[1]+'\n')
else: # negated alias
f.write('1 -%d 0 -1 # ' % cnt_1 + name[1] +'\n')
else:
if aliases[i][1] == 0: # noalias
cnt_2 = cnt_2 + 1
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] +'\n')
elif aliases[i][1] == 1: # alias
f.write('2 %d 0 -1 # ' % cnt_2 + name[1] +'\n')
else: #neg alias
f.write('2 -%d 0 -1 # ' % cnt_2 + name[1] +'\n')
f.write('\n')
sc = model.jmimodel.get_variable_scaling_factors()
z = model.z
rescale = (model.get_scaling_method() ==
jmi.JMI_SCALING_VARIABLES) and (not scaled)
# Write data
# Write data set 1
f.write('float data_1(%d,%d)\n' % (2, n_parameters + 1))
f.write("%.14E" % data[0,0])
str_text = ''
for ref in range(n_parameters):
#print ref
if rescale:
#print z[ref]*sc[ref]
#print "hej"
str_text += " %.14E" % (z[ref]*sc[ref])
else:
#print z[ref]
#print "hopp"
str_text += " %.14E" % (z[ref])
f.write(str_text)
f.write('\n')
f.write("%.14E" % data[-1,0])
f.write(str_text)
f.write('\n\n')
# Write data set 2
n_vars = len(data[0,:])
n_points = len(data[:,0])
f.write('float data_2(%d,%d)\n' % (n_points, n_vars))
for i in range(n_points):
str = ''
for ref in range(n_vars):
if ref==0: # Don't scale time
str = str + (" %.14E" % data[i,ref])
else:
if rescale:
str = str + (" %.14E" % (data[i,ref]*sc[ref-1+n_parameters]))
else:
str = str + (" %.14E" % data[i,ref])
f.write(str+'\n')
f.write('\n')
f.close()
else:
raise Error('Export on binary Dymola result files not yet supported.')